论文部分内容阅读
如何识别虚假财务报告是一件十分困难的事情。针对虚假财务报告的特点,设计一个基于支持向量机的虚假财务报告识别模型将有助于该问题的解决。根据1999-2002年的年度审计报告意见类型,从上市公司中选取44家虚假财务报告样本,并按照一定的标准选择了44家对比样本,这88个样本构成训练数据集。类似地,从2003-2006年的上市公司中,选择了73家虚假财务报告样本和99家对比样本,这172个样本构成测试数据集。我们使用训练数据集对支持向量机模型进行训练,并将训练后的模型对测试数据集进行测试,取得了较好的实验结果。
How to identify fake financial reports is a very difficult thing. In view of the characteristics of false financial reports, designing a false financial report recognition model based on SVM will help to solve this problem. According to the type of opinion of the annual audit report of 1999-2002, 44 samples of fake financial reports were selected from the listed companies and 44 comparative samples were selected according to certain criteria, and the 88 samples constituted the training data set. Similarly, from 2003-2006 listed companies, 73 false financial report samples and 99 comparative samples were selected, and these 172 samples constitute the test data set. We use the training data set to train the SVM model, and test the test data set after the training model, and get better experimental results.